[IMPALA-9632] Implement ds_hll_sketch() and ds_hll_estimate() functions

• Type: New Feature
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Component/s: Backend, Frontend
• Epic Link: unlabelled-IMPALA-9593
• Epic Color: ghx-label-12
• Assignee: Gabor Kaszab
• Reporter: Gabor Kaszab
• Created: 09/Apr/20 12:00
• Updated: 07/Jul/20 21:06
• Resolved: 07/Jul/20 14:12

Description

These should be built-in functions that use DataSketches functionality that was integrated by IMPALA-9631.

ds_hll_sketch() should receive a primitive expression and return a sketch.
ds_hll_estimate() should receive a sketch and return a primitive that is the cardinality estimate for that set of data provided to the sketch.

Usage:
select ds_hll_estimate(ds_hll_sketch(col_name)) from table_name;
Returns a cardinality estimate (similarly to ndv() ) for that particular column.

Hive change that introduced the same: https://issues.apache.org/jira/browse/HIVE-22940